Sunday, August 14, 2011

Using Field Experiments To Diagnose the Causes of Our Economic Malaise

Have you noticed that economists do not agree about what are the causes and cures for our current malaise?   Right now we seek to reduce the budget deficit and to reduce unemployment.   If more people worked and paid taxes, then unemployment insurance claims would be lower and the IRS would collect more tax revenue, so it appears that job growth would address both of these issues.  

Firms hire workers when the marginal value product is greater than the wage they must pay.  During this time of macroeconomic uncertainty and uncertainty among academic economists concerning what is the disease and what is the correct medicine to order, I suggest that it is time to experiment and follow the advice of Banerjee and Duflo.

To quote these authors;  "Perhaps even more importantly, they are often in a position to midwife the process of policy discovery, based on the interplay of theory and experimental research. It is this process of “creative experimentation”, where policymakers and researchers work together to think out of the box and learn from
successes and failures, that is the most valuable contribution of the recent surge in experimental work in economics."

So, permit me to propose some randomized experiments;

Experiment #1;   Take the counties in the United States where unemployment is over 11% right now and randomly assign them to treatment and control.  In the treatment counties, allow any employer in these counties to be exempt from minimum wage laws and from unionization rules for 24 months.   In this experiment, the job creation in these counties will be compared to the control counties in a simple "before/after" comparison.  In the control counties, make no changes to existing labor regulations.

Experiment #2;   For the same subset of high unemployment counties, randomly assign counties to "health insurance" categories.  For firms that locate in counties assigned to the "treatment group", these firms will have to pay for no health care insurance for workers they hire. Instead, the government will provide health care offering Medicare coverage for the hired workers.

These experiments will help to settle the issue of whether labor market compensation is reducing job creation. If firms in the treated counties under experiment #1 and #2 continue to refuse to hire workers, then this is strong evidence that U.S workers lack the skills to be competitive in this global economy and we face a long run problem.